Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-developers-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-developers/>
List-Post: <mailto:cygwin-developers AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-developers-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-developers-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin-developers AT sources DOT redhat DOT com
Date: Thu, 6 Sep 2001 11:04:34 +0400
From: egor duda <deo AT logos-m DOT ru>
X-Mailer: The Bat! (v1.53 RC/4)
Reply-To: egor duda <cygwin-developers AT cygwin DOT com>
Organization: deo
X-Priority: 3 (Normal)
Message-ID: <4463552103.20010906110434@logos-m.ru>
To: Charles Wilson <cwilson AT ece DOT gatech DOT edu>
CC: cygwin-developers AT cygwin DOT com
Subject: Re: jik, cwilson, earnie -- show stoppers
In-Reply-To: <3B96991D.6060107@ece.gatech.edu>
References: <20010905165512 DOT A3778 AT redhat DOT com> <20010905165709 DOT A3907 AT redhat DOT com>
 <3B96991D DOT 6060107 AT ece DOT gatech DOT edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi!

Thursday, 06 September, 2001 Charles Wilson cwilson AT ece DOT gatech DOT edu wrote:

CW> I'm at a total loss, Chris. I can no longer reproduce "my" error -- EVEN 
CW> when using the SAME cygwin1.dll that crashed before, using the same 
CW> sequence of events that previously *always* triggered a crash (at some 
CW> random point during the build).  Does that mean the bug is fixed, in the 
CW> absence of any patches?  Of course not.  But I can't help debug it 
CW> anymore, because I have nothing to debug!

CW> Now, fortunately for all of us, Egor (somehow) seems to have reproduced 
CW> "my" problem on his machine, and supplied a patch that seems to fix it. 
CW>   On his machine.  If the bug he is seeing is the same as the one I used 
CW> to see, but don't see any longer.  I dunno if Egor's patch fixes *my* 
CW> bug on *my* machine, since my machine magically fixed itself.  Without 
CW> changing anything.

CW> Bleah.  (I'm a bit frustrated.  Can you tell?)

CW> I'm willing to do what I can to help with this -- but how CAN I?  The 
CW> bug dissappeared!  (Seriously, I'm open to suggestions....)

as i mentioned in cygwin-patches@, the problem was that buffer for
security descriptor was only 256 bytes while actual descriptor was 268
bytes (and it triggers only at file creation!). this situation is very
sensitive to the layout of your file system, permission propagation
bits etc., which could be easily disappear as your file system change.

Btw, i've found it while testing several new tests for testsuite :).
Unfortunately, the test that failed was test for select() and it
seems that it's yet another pure coincidence that it triggered this
bug. I'll check new tests in today, and it'd be interesting if someone
else could reproduce this bug as well.

Egor.            mailto:deo AT logos-m DOT ru ICQ 5165414 FidoNet 2:5020/496.19